233.7 ± 0.2
226.9 ± 0.1
281.2 ± 0.4
191.2 ± 0.1
183.2 ± 0.1
314.4 ± 0.5
191.8 ± 0.1
299.4 ± 0.3
248.8 ± 0.1
179.3 ± 0.1
265.3 ± 0.3
278.0 ± 0.2
306.9 ± 0.2
285.5 ± 0.2
194.8 ± 0.1
190.1 ± 0.1
212.0 ± 0.1
168.3 ± 0.1
249.3 ± 0.2
179.8 ± 0.1
364.5 ± 0.8
202.5 ± 0.1
241.1 ± 0.2
281.3 ± 0.2
199.6 ± 0.1
222.9 ± 0.2
233.1 ± 0.1
201.9 ± 0.1
170.1 ± 0.1
226.4 ± 0.2
262.7 ± 0.2
181.9 ± 0.1
240.5 ± 0.1
225.0 ± 0.2
232.8 ± 0.1
177.9 ± 0.1
216.1 ± 0.1
255.4 ± 0.1
232.4 ± 0.1
194.2 ± 0.1

299.5 ± 0.2
289.8 ± 0.1
361.8 ± 0.2
233.4 ± 0.1
212.1 ± 0.1
413.2 ± 0.2
229.9 ± 0.1
394.9 ± 0.2
327.2 ± 0.1
208.5 ± 0.1
338.9 ± 0.2
363.1 ± 0.2
423.2 ± 0.2
379.9 ± 0.2
232.0 ± 0.1
232.0 ± 0.1
266.2 ± 0.1
193.6 ± 0.1
325.5 ± 0.2
211.2 ± 0.1
488.5 ± 0.5
252.7 ± 0.1
312.9 ± 0.2
381.4 ± 0.2
238.0 ± 0.1
285.0 ± 0.2
300.9 ± 0.1
254.9 ± 0.1
196.5 ± 0.1
291.0 ± 0.2
354.3 ± 0.2
215.3 ± 0.1
314.9 ± 0.2
276.6 ± 0.1
298.3 ± 0.1
212.6 ± 0.1
271.7 ± 0.1
337.2 ± 0.1
297.0 ± 0.2
238.6 ± 0.1

Input mol2 files with partial charges are available as a separate file in the supporting information, as are AMBER formatted parameter and
coordinate files.
Table 5 shows the application of Student’s t-test to comparing the calculated and experimental means by each functional group. For almost
every functional group, the test indicates that the means are significantly different. This is not surprising since we observed a mean error of
0.676±0.002 kcal/mol – solutes systematically prefer the gas phase too much relative to experiment.

26


